Development of innovative methodologies for non-invasive characterization of metal artefacts of archaeological, historical, and industrial interest, through neutron diffraction and neutron imaging techniques

(1)

DOTTORATO DI RICERCA IN

Scienza per la Conservazione dei Beni Culturali

CICLO XXVI

COORDINATORE Prof. Piero Baglioni

Development of innovative methodologies for non-invasive characterization of metal artefacts of

archaeological, historical, and industrial interest,

through neutron diffraction and neutron imaging techniques

Settore Scientifico Disciplinare FIS/07

Dottorando Tutore

Dott. Salvemini Filomena, Floriana Prof. Zoppi Marco

Co-Tutore Prof. Grazzi Francesco

Coordinatore Prof. Baglioni Piero

(2)

(3)

Abstract

Metal artefacts of archaeological and historical interest represent a large fraction of the rich cultural heritage which is conserved in national and private museums worldwide. As such, these objects, when properly studied, can reveal secrets of the past human history and we are only allowed to treat them using the most delicate care and avoiding, as much as possible, to accelerate the natural and unavoidable aging process. Moreover, as the technological knowledge evolves, we are not allowed to investigate these objects using invasive tools that might prevent further more involved techniques to operate properly on the same objects by future generations.

Until recently, the research activity in this field was mainly based on standard point-based analytical techniques like, for example, metallography. Traditional analysis, however, though extremely effective, is not always suitable for rare and unique objects of high scientific, cultural, and economic value. Lots of interesting metal items fall in this category: sculptures and artistic artefacts, archaeological finds, rare coins, and even meteorites. It has been shown that this technique, which is mainly based on a careful surface analysis following a cleaning and preparation process, can be effectively complemented by a well-recognized, non-invasive experimental approach, based on the use of thermal neutrons.

Thanks to their high penetration power, neutrons represent an ideal tool to probe the microscopic properties of bulk dense materials and can be used to characterize the microscopic structure (at the atomic level) of metal artefacts. As a matter of fact, neutron techniques are used to determine the qualitative and quantitative presence of different phases, as well as the presence and distribution of texture and residual strain at the atomic level. From this wealth of data it is possible to obtain information on the conservation status of the artefact, as well as on the smelting and smithing procedures, through identification of some peculiar signatures related to these processes. In fact, the study of forging techniques, and their time-evolution, represents one of the most interesting topics in the investigation of the manufacturing methods, which are considerably different among various cultures. To this aim, neutron diffraction can be used to quantitatively determine the relative amount of the various crystal phases composing the artefact, its conservation status, the presence of texture, measurements of residual strains, and the determination of the grain size of the phases in the sample.

In addition, neutron imaging techniques allows obtaining bulk detailed information on the micro-structural properties of the samples, through the tomographic reconstruction of the object’s macroscopic cross section. However, this technique can be further improved, to achieve materials discrimination, by a proper selection of the neutron energy. In fact, monochromatic neutron beams give the possibility of modifying the image contrast for different phases taking advantage of the abrupt change of the attenuation coefficients in the proximity of the so-called Bragg cut-off.

In the present work, we have developed a comprehensive study of the metallurgy of museum metal artefacts of historical and archaeological interest aiming to investigate composition, assembly methods, and structural variations pertaining to different cultures and historical environments. The study has been developed in cooperation with several Museum Institutions (like, for example: Museo Stibbert – Firenze (IT) and The Wallace Collection - London (UK)) and some private collectors.

(4)

Beyond time-of-flight neutron diffraction, which represents the workhorse of the neutron techniques, white beam and energy selective radiography, tomography, and laminography have been successfully applied to determine the material composition and microscopic structures of the analysed samples. From these results, we could obtain information on the smithing techniques and the working process of the studied of ancient artefacts. For this investigation several experimental European facilities have been used, like the ICON, NEUTRA and POLDI beam-lines at the Paul Scherrer Institut (CH), the CONRAD II beam-line at the Helmholtz-Zentrum Berlin (DE), and the INES and ENGIN-X beam-lines at ISIS (UK).

Japanese sword (katana) blades pertaining to the Koto (987–1596) and the Shinto (1596– 1781) periods in Japan have been characterized using neutron diffraction and neutron-imaging techniques (in this case both white beam and energy selective neutron-imaging have been used). However not only the blades were the object of the present investigation, but also the composite structures of the metal components at the handle’s extremities of Japanese swords have been investigated: the hand-guard tsuba, the hilt collar fuchi, the pommel

kashira, the small knife-handle fitting into the scabbard kozuka. In addition, we extended

the study to the inner metal structure and manufacturing techniques of kabuto (helmets), which represent beautiful examples of past technology made in Japan in the 17th Century. The present investigation, however, was not limited to the study of the Japanese historical objects but was extended to other cultural environments. Thus, we have done experiments on a set of five kris,the distinctive weapon of Malaysia and Indonesia, and on a kanjar, the Indian dagger. Morphological information and material distribution in the whole volume of these objects allowed us to identify the working procedures and, when possible, to define the authenticity of the investigated samples.

The study has been enriched by a methodological test, aiming to demonstrate the equivalence of the information content between the classical (invasive) technique of metallography and neutron diffraction and imaging techniques (non-invasive). An enlightening comparison could be obtained studying the cross sections of a Toledo-like sword, which has been studied thought neutron imaging. Fragments, taken at different height of two sacrificed samples, were investigated through energy selective images and the results have been compared to those obtained from standard metallography.

Once the neutron techniques were validated, on a qualitative and quantitative basis, a massive investigation was conducted on a large number (91) of styli, the small bar made of either iron, bronze or bone, used by the Romans for writing onto wax-coated surface of wooden flat tablets. All these samples, coming from the same archeological site of Julia Concordia - Venice (Italy), were carefully analyzed by neutron techniques and have disclosed peculiar morphological features related to the working techniques of ancient Romans.

Finally, it is worthwhile to note that, beyond the standard neutron diffraction and imaging experiments, a relevant number of samples (two selected Japanese swords and 11 Roman

styli) that have been analyzed to obtain the residual strain mapping aiming to obtain

(5)

1 Introduction

An accurate knowledge of the composition (phase distribution) and the microstructure (size, shape, and orientation of grains) of materials represents the equivalent of the Rosetta

stone in material science. Until recently, especially concerning metal objects, this task was

mainly fulfilled basing on standard analytical techniques like, for example, metallography. Traditional analysis, however, though very accurate, is not always suitable for rare and unique objects of high scientific and economic value. Lots of interesting items related to cultural heritage fall in this category: sculptures and artistic artefacts, archaeological finds, rare coins, and even meteorites. It has been shown that this well established technique, which is mainly based on surface analysis, can be effectively complemented by an emerging non-invasive experimental approach based on thermal neutron analysis. In particular, neutron-imaging techniques, in combination with neutron diffraction, provide an attractive approach to determine the morphology, the qualitative and quantitative presence of different phases, as well as the presence and distribution of texture and residual strain. From this wealth of data it is possible to obtain information on the conservation status of the artefact, as well as the smelting and smithing procedures, through identification of some peculiar signatures related to these processes.

In order to characterize different examples of Cultural Heritage objects, disclosing different open questions about their origin or production method, this study, devoted to analyze the potential of the aforementioned neutron techniques, was developed in cooperation with several Research Groups and Museum Institutions.

In this thesis neutron diffraction and imaging techniques are applied for the development of methodological approaches aiming to a non-invasive characterization of metal artifacts of historical, archaeological and industrial interest pertaining to different periods and cultures.

Chapter 2 is dedicated to ancient metallurgy; briefly travelling through the development of human technology in time, space and culture. Here, the research questions that led to the present investigation are introduced. In Chapter 3, we briefly recall a number of introductory concepts concerning the study and technology of metals: alloys, working techniques, meaning of stress, strain and texture and so on. In Chapter 4, we introduce the neutron techniques with a particular attention to neutron imaging and diffraction used as diagnostic techniques for the characterization of metals. Finally, Chapter 5 reports a panoramic view of the various measurements and the results obtained for the different ancient and historical metal objects.

(8)

Chapter 2

(9)

2 Ancient metallurgy

Slowly moving from the Stone-age to a more advanced stage of mankind, our forefathers had several occasions to meet different kind of metals that gave, once the relative technology was developed, a sensible impulse to the human progress. Initially, of course, only those metals that were found as native ones (e.g. copper, silver, and gold) were used for the production of tools, weapons, and jewels. These metals were “easily” collected, and worked, to be used in the same areas where they could be found or to become precious objects of exchange. Some time, also meteoric iron was found, albeit in small and usually rapidly exhausted quantities. However, in this case, the technological skills necessary to transform this material into a useful tool were delayed until the copper technology reached its climax [Buchwald 2005].

Native copper was the first metal that was extensively used in Asia, Europe and North America [Coghlan 1975, Tylecote 1987, Wayman 1989]. It could be found in the lower part of the gossan, the secondary oxidized minerals which then covered many sulphidic copper deposits. The early metallurgy of native copper has been widely discussed by historians and metallurgists. Early evidences were found in southeastern Turkey around the end of the 9th_{millennium B.C. and the beginning of the 8}th_{[Maddin et al. 1999]. It appears}

plausible that once early craftsmen had discovered the possibility to use charcoal fire to heat native copper in a clay crucible, up to the melting point (1083°C), it was not a long step to discover other means of pure copper production, starting from the attractive colorful oxycabonates, malachite and azurite. By this technique, a mixture of metal and slag2.1 was formed in the furnace, and this was subsequently broken up and melted in crucible to purify the metal component [Buchwald 2005]. The subsequent discovery that adding a different metallic ore to the melt would improve the best technological features of pure copper [Wheeler and Maddin 1980], lead to the production of arsenical bronze. This was obtained from co-smelting2.2_{of ore minerals containing Cu and As. At a later stage, the}

exploitation of other metal ore mixtures, such as those containing tin, could lead to the formation of an alloy (a solid solution of two or more metals), such as bronze [Williams 2003].

Apparently the development in copper technology took place independently in many different places, as the early locations appear very widely spread. Evidences of these early attempts, dating back to the 6th_{millennium B.C., have been documented by archaeologists}

operating in Central Anatolia, in northern Mesopotamia, and in northwestern Iran. The first smelting activities, appearing in the Middle East, start from the 4th_{millennium B.C. [Flawn}

1966, Hauptmann 1985, Hauptmann et al. 1992]. At any rate, it is customary to attribute the beginning of metallurgy to the early discovery of smelting and extracting copper from its ores [Buchwald 2005].

2.1_{A slag is defined as the}_by-product_of_smelting_ore_{to purify}_metals_{. In nature, the ores of metals are found}

in impure states: an intimate association of metal (sulphide, oxide, native state), and an undesired embedding rock (gangue) known as tout venant [Yalçin and Pernicka 1999]. A slag generally consists in a mixture of metal oxides and its composition can be influenced by several factors. They are: nature of the ores and

gangue, charcoal and wood ashes, furnace construction material (including lining, bricks, tuyeres, temper

added to clays used as furnace linings, etc.), fluxes added, process condition (heat distribution, air intake, furnace profile and height, retention time of slag within furnace), cooling condition [Benvenuti 2004].

2.2_{During smelting, the ore is heated beyond the melting point, ordinarily in the presence of reducing agents}

(charcoal or coke). Thus a fuse composed by silicates is separate from the metal fraction, which forms as a solid in it. During the cooling of the fuse, impurities are separated from the metal and can be easily removed [Backmann 1968].

(10)

In particular, the Sumerians were the first to adopt tin bronze technology and the metallurgy of the Early Dynastic period (around 3000 B.C.) was well developed [Moorey 1994]. Despite the lack in ores, acquired from trade with the Near East, Mesopotamia was extremely rick in natural fuel. The advanced use and knowledge of the fuels and their properties made the Sumerian technology of bronze decidedly superior respect to contemporary cultures. Extending from India to Anatolia, the bronze metallurgy was spread spasmodically, appearing in the Indus valley in about 2500 B.C. and progressing westwards through Europe from about 2000 [Kaufman 2011].

Unlike copper, iron never appears in nature, in its native form. Iron ores are very widespread, but the extraction of the metal component is not simple, due to the high melting-point temperature at about 1500°C. On the other hand, the Iron Age could not develop until successful techniques for the reduction of iron ores had been devised and disseminated.

It is very likely that some attempts of iron working could have been attempted using the copper technology and meteoric iron. However, the erratic availability of the base material and the substantially inadequacy of the technique did not allow a regular production of iron objects, even though some findings of iron objects were exceptionally recovered in sites unequivocally attributed to the copper age [Grazzi 2012]. In addition, due the technology limitations, the quality of the first iron smelting attempts was not of immediate advantage to the Ancient World. Iron smelting seems to have been first developed, with some regularity and continuity, somewhere between the Caucasus and the Fertile Crescent early between the 3rd_{and the 2}nd_{millennium B.C [Forbes 1964, Scott 1991].}

Following the destruction of the Hittite Empire, in the 1200 B.C., the knowledge of iron-making spread fairly quickly around the Near East and it was exploited in a considerable scale by the Assyrian, which made a large scale use of iron. Greeks too used iron tools quite extensively, although they continued to employ bronze swords and armors. As a matter of fact, the Trojan war, also dated around 1200 B.C. was apparently fought using bronze weapons. According to some historians [e.g. Herodotus], Tyrrhenian populations, migrating from southern Anatolia around 12th_{XII century B.C., diffused the iron}

metallurgy in the Mediterranean area giving rise, likely, to the Etruscan civilization [Pallottino 1998]. In the northern Europe, north of the Alps, the westward movement of Celtic populations spread the knowledge of iron weapons and tools after the 5th century B.C. [Cleere and Scott 1984, Tylecote 1987].

The very low-carbon iron produced in the bloomery hearth furnaces, where bellows were used to force air through a pile of iron ore and burning charcoal, was inferior in hardness, as well as in corrosion resistance, to copper alloys. The success and upsurge of iron depended entirely on its ability to become carburized2.3_{and converted onto steel.}

It has also been suggested that the transition to iron was due to shortage in the supplies of copper and, especially, tin. In addition, there was a fuel-saving incentive to change from bronze to iron and, moreover, the iron ore was much more widespread [Buchwald 2005]. Anyway, iron weapons and armors did not become superior to bronze until the fundamental discovery that quenching2.4, after carburization, resulted in a dramatic increase in hardness. The process is difficult to manipulate, since the hardness is due to the formation of martensite2.5_{, and an excess of this phase leads to embrittlements (and}

fragility) of the tool.

2.3_{Carburization is the process of increasing the carbon content of a metal.} 2.4_{The act of quickly cooling a metal or alloy by plunging into cold water or oil.}

2.5_{A non-equilibrium metastable phase resulting from the transformation, without diffusion, of austenite (see}

(11)

Quenching is mentioned by Homer (perhaps in the 10th_{or 9}th_{century B.C.) but the}

difficulties of controlling the carbon content in steel meant that quenching remained a hit-and-miss process, and therefore avoided by many smiths, for a long time to come.

The Roman army initially organized itself on the Greek and Macedonian models, also in terms of equipment, but contacts with the Celts and the experience of the Punic Wars led them to adopt the Celtic sword-making technology [Williams 2003].

In Central Europe ironworking developed within the Celtic sphere of influence from the 8th

to the 7th century B.C. during a period that can hardly be determined. Iron soon became an important factor in everyday’s life. The pulse for the development of iron-making in the Celtic area presumably moved from the Mediterranean and Near East towards Central Europe and could have taken several routes over the Alps and towards the western and eastern countries of southern Europe [Pleiner 1996]. According to the current school of thought, iron, in Europe, was exclusively produced in direct process before the blast furnace was developed at the end of the Medieval period. In this phase of the technological progress, iron was made in small furnaces, through the reduction of the ore into a pasty state, because temperature was not sufficiently high to melt iron [Shurmann 1958, Gassmann 1998]. In current use, a charge of ore and charcoal was placed in the furnace and, during the smelting process, some of the lighter impurities, i.e. the melted ones, were tapped off as slag until, at the end, cinders, slag and the so-called bloom2.6_{of iron remained}

at the bottom of the furnace. Repeated heating and forging of the bloom would be necessary to expel much of the slag and consolidate the metal part to produce a bar of wrought iron [Goodall 2012]. Then, the obtained product could be carburized. The process of carburizing wrought iron consisted in taking an iron or low carbon steel piece, packing it with carbonaceous material in a sealed vessel, and then baking at high temperature (at red heat) for a long period of time. Thus carbon would leech into the metal creating steel, eventually [Johnson 1999]. Pieces of iron (low carbon content, C<0.2 wt% C) and steel (higher carbon content, C>0.2 wt% C) could be welded together (by hot hammering) to form a sword, which was then hardened by some form of quenching. This procedure was adopted in both European and East Asian smithing techniques. In Europe, it led to the development of the piled and the pattern-welded sword. The subsequent development of large bloomeries2.7_{enabled to make steel in much larger quantities, so that the later}

European Middle Ages (14th century) saw the development of suits of steel plate armor as well as all-steel swords. In Japan this process was brought to its highest level, where it formed the basis of sword-making, almost until modern times [Williams 2003].

The date and origin of the introduction of iron artifacts and iron working into India remains a much debated problem, connected with the equally debatable question of its association with the supposed arrival, in the 2nd_{millennium B.C., of immigrants from the west. It is}

evident that iron first appeared in India between 1300 and 1200 B.C.. However, it is generally agreed that the beginning of the 1st_{millennium B.C. saw the advent of iron in}

India during a period of peaceful settlements and political stability (the Maurya period)2.8

[Prakash and Tripathi 1986].

India was famed in literary and history accounts since Greek and Roman time for the traditional crucible steel. According to Will Durant, the technology passed to the Persians

2.6_{A roughly finished metallic product; specifically, the spongy mass of iron produced in a bloomery furnace}

in which the iron is reduced in situ and is not molten during reduction. The crude iron bloom must be extensively worked at red heat to consolidate the iron and remove excess slag and charcoal [Prakash and Tripathi 1986].

2.7_{The furnace for the extraction of iron from the ore.}

2.8_{The Maurya Empire was a historical power in ancient India, ruled by the Mauryan dynasty from 322 to}

(12)

and from them to Arabs who spread it through the Middle East. In the 16th century, the Dutch carried the technology from South India to Europe, where it gave ride to steel mass-production [Porter 2003].

In this process, small pieces of iron can be separated from the bloom and then heated in a close crucible together with charcoal (which is almost entirely composed by carbon), until a partial or total melting takes place. Rapid absorption of carbon can lead to the formation of cast steel (“crucible” steel), with a very high (1.2-1.6 wt% C) carbon content, which needs further little hardening. Some steel of this type seems to have found its way into Europe during the early Middle Ages. Controlled cooling and forging can then develop a pattern, resembling watered silk, on the surface of the blade (wootz steel, misnamed

Damascus steel). This was the procedure used in Iran, Central Asia and India, where it

remained in operation until 19th_{century, with products that were high in quality but small}

in production scale [Williams 1963, Panseri 1965, Craddock 1995].

In China, bloomery-based ironworking was probably introduced through Central Asia [Pigott 1999]. However, from the 5th century B.C., a liquid product of the furnace was more usual. Such a furnace was called a blast furnace and differs from the bloomering only in a somewhat larger size and a different fuel over ore ratio. Owing to the sensible decrease of steel melting-point with increasing carbon content, a mixture over 2% carbon will melt at 1150°C, forming a liquid called cast iron or pig iron. This material can be cast into shape, but cannot be forged, nor hardened by quenching, and would seem useless for making weapon. Therefore, cast iron was converted into wrought iron by decarburization; to this aim, the steel was reheated and oxidized in air for several days in the finery. This technology, called the indirect iron process, was locally in use since the 1st_{century B.C.;}

The spread of this technique was favoured by the considerable scale economy it implied, allowing to work continuously, with a considerable saving in fuel per work unit, and avoiding the need to demolish the furnace for the extraction of the metal. Thus, large-scale production was favoured, while the bloomery gradually became uncompetitive, [Williams 2012, Craddock 1995].

In the late 12th_{century A.D. the indirect iron process became known in northern Europe,}

through Sweden, but it was spread, and commonly used in the resto of Europe, only during the 15th_{century A.D.. On the other hand, many scholars believe that blast furnaces} _have

developed independently in Western Europe andChina, albeit many centuries earlier in the latter[Buchwald 2005].

To be mentioned is the unique technology developed independently in Japan, where the traditional Japanese furnace, known as a tatara (Fig. 2.2), was developed. This is a hybrid type of furnace whose origin is still debated [Juleff 2009, Inoue 2009]. The tatara does not

appear to be a Japanese invention and it may have come from Manchuria, by way of the Korean peninsula in the 6th_{or 7}th_{century A.D.. By the 9}th_{century, the tatara was in use}

throughout Japan, primarily by small groups of steelworkers, mining and smelting their own steel, and later evolving in mass production in selected centers [K. Nagayama 1997]. The tatara incorporated bellows, like the European blast furnace, but was constructed of clay; this furnace would eventually collapse after the first use [Wittner 2007]. According to existing archeological records, the first tatara were built during the middle part of the 6th

century A.D. [Juleff 2009]. Due to the large scale of the tatara, as compared to its European, Indian and Chinese counterparts, the temperature at a given point would vary depending on the height in the furnace. Therefore, different types of iron could be found (with varying degrees of carbon content), at different heights, inside the furnace, ranging from wrought iron at the top of the tatara (furthest from the heat, lowest temperature), to cast iron towards the middle, and finally steel towards the bottom [F. Grazzi 2011]. Importantly, the tatara does not exceed 1500°C, so that it is unable to completely melt the

(13)

iron. Obviously, the metal-workers could understand the differences between the various types of iron and they separated out and selected different portions of the “bloom” accordingly [Inoue 2009]. 2.9 _{From the early introduction of metal tools by the nearby Asia}

continents, during the Yayoi-period, the Japanese smiths evolved acquiring sophisticate skills that culminated in the forging of weapons, particularly swords [K. Nagayama 1997].

2.9_{The metallurgy of others metals such as gold, silver, zinc, mercury and lead was avoided because out of}

the main themes discussed in this work.

Figure 2.2: Cross section of a typical tatara. The complex structure below the ground is for the

insulation, damp-proofing, and drainage. The clay driers are hollow; wood was burned in them at the time of construction [http://www.thejapanesesword.com/tatara].

(14)

Chapter 3

(15)

3 Materials

3.1 Metals

Materials in this group are composed of one or more elements (e.g., iron, copper, tin, zinc, silver, gold, nickel, ...), and often also non-metallic elements (e.g., carbon, nitrogen, oxygen, ...) in a relatively small concentration3.1_{. Atoms in metals and their alloys are}

arranged in very defined crystal structures and are relatively dense. Concerning the mechanical characteristics, these materials are relatively stiff and strong, yet are ductile and are generally resistant to fracture.

As is well known, metals contain large numbers of non-localized electrons, which are not bound to particular atoms. Many properties of metals are directly attributable to these electrons. For example, metals are extremely good conductors of electricity and heat, and are not transparent to visible light; a polished metal surface has a lustrous appearance. In addition, some metals (i.e., Fe, Co, and Ni) have desirable magnetic properties [Callister 2009].

3.2 Crystal structure

Solid materials may be classified according to the regularity with which atoms, or ions, are arranged with respect to one another. The distinct feature of a crystalline material is the placement of its atoms in a periodic array over large atomic distances, giving rise to long-range order where the atoms will position themselves in a repetitive three-dimensional pattern and each atom is bound to its nearest-neighbors. Under normal conditions, all metals form ordered and symmetrical arrangements called crystal lattice. Some of the properties of crystalline solids depend on the microscopic structure of the material, and therefore from the way in which atoms, ions, or molecules are spatially arranged [Callister 2009].

Every crystalline structure describing the regular position of atoms in a crystalline material may be considered in terms of a lattice of points in real space. Associated with each point is an array of atoms called basis. Thus, a structure comprises a lattice with its basis. The lattice of points can be described in terms of lattice vectors in real space, with every point on the lattice l given in terms of the fundamental translation vectors, or crystal axes, a, b, c, which are chosen so that:

L=n1a+n2b+n3c (3.1)

where n1, n2, n3 are integers. The lattice points can thus be said to have 3D translational

symmetry. The values of a, b, c are called the lattice constants, which define the lattice together with the angles between these axis  and , where is the angle between b and

c and so on.

Associated with lattice is the concept of unit cell, a building block that, by repetition through linear translations, covers all points on the crystal lattice. The smallest unit cell is called the primitive unit cell, usually chosen to have one lattice point at each corner; the lattice points will then lie on what is called a primitive lattice [Hutchings et al. 2005]. The Bravais lattice is the set of all equivalent atoms in a crystal that is brought back onto itself when the whole set is displaced by the length of an integer value of a unit vector in a direction parallel to a unit vector [Kossevich 1999].

(16)

In the three-dimensional space, the Bravais lattices are classified in a total of 14 possible lattices, according to the symmetry of rotations and reflections, and to the location of the lattice site [Kittel 1996].

Cubic and hexagonal lattices are of relevant interest here, since a large number of materials, especially metals, have this kind of lattice. Three relatively simple Bravais lattice are found for most of the common metals: face-centered cubic (fcc), body-centered cubic (bcc), and hexagonal close-packed (hcp), which, incidentally, represent the three most dense lattice configurations [Kittel 1996].

The crystal structure found for many metals has a unit cell of cubic geometry, with atoms located on each of the corners and on the centers of all the cube faces (fcc). The atomic packing factor3.2_{(APF) for this structure, is 0.74, which is the maximum possible packing}

value for spheres of the same diameter. The coordination number, i.e. the number of nearest neighbors, is 12 that is the same for fcc and hcp lattices. Metals typically have relatively large atomic packing factors, to maximize the shielding provided by the free electron cloud. Some of the most familiar metals, having this crystal structure, are copper, silver, and gold (Fig. 3.1-a).

Another common metal crystal structure is the bcc where the atoms are located on the corners and the center of a cubic cell. The coordination number this structure is 8 and, correspondingly, the atomic packing factor is 0.68. Iron, tungsten, and chromium exhibit a

bcc structure (Fig. 3.1-b).

The hcp crystal structure, another close-packed structure like fcc, has a hexagonal unit cell (Figure 3.1-c). The coordination number and the atomic packing factor for this structure are the same as for fcc: i.e. 12 and 0.74, respectively. The HCP metals include cadmium, magnesium, titanium, and zinc [Callister 2009].

A b c

Figure 3.1:The reduced-sphere unit cell models for a the face-centered cubic crystal structure, b the body-centered

cubic crystal structure, c hexagonal close-packed crystal structure [Callister 2009]. 3.2.1 Cristallographic planes and directions

The orientation of crystallographic points, directions, and planes are specified in terms of indexing schemes. The basis for determining index values is the unit cell, with a right-handed coordinate system consisting of three axes (x, y, and z) situated at one of the corners and coinciding with the unit cell edges [Callister 2009].

The application of a set of rules leads to the assignment of the Miller indexes (hkl). The

Miller indexes are conventionally defined as the reciprocal multiples of the axis intercepts,

reduced to the smallest integers having the same ratios.

3.2_{The APF is the sum of the sphere volumes of all atoms, bordering one another, within a unit cell divided}

(17)

In the single crystal, the periodic and repeated atomic pattern extends throughout the entire crystalline solid without interruption. However, this condition is very rare in real metals that, instead, are characterized by a generally polycrystalline structure composed by a collection of single crystal grains arranged together.

3.3 Alloys

A metal alloy is a solid mixture composed of two or more elements where the main component is always a metal. Most familiar alloys are those where impurity atoms have been added intentionally to impart specific characteristics to the material and to improve its properties. The addition of an impurity to a metal will result in the formation of a uniform solid solution and/or formation of a new phase, depending on the kinds of impurity, their concentrations, and the temperature of the alloy.

A solid solution may form when the impurity added atoms occupy substitutional or interstitial sites and are uniformly and randomly dispersed within the solid. In this case, the original host crystal structure is maintained and no new phase is formed, giving rise to a compositionally homogeneous resulting material. Several features determine the composition of substitutional solid solution: the atomic diameters, and electronegativities, should be similar; both elements should have the same crystal structure (or very similar ones, like fcc and hcp); the number of valence electrons in the impurity atoms should be equal, or smaller, than the host material [Pearson 1964]. In the particular case of interstitial solid solutions, impurity atoms fill the voids or interstices among the host atoms. For metallic materials that have relatively high atomic packing factors, these interstitial positions are relatively small. Consequently, the atomic diameter of an interstitial impurity must be substantially smaller than that of the host atoms. Normally, the maximum allowable concentration of interstitial impurity atoms is low (less than 10%). Since even very small impurity atoms are ordinarily larger than the interstitial sites, they introduce some lattice strains on the lattice of the host atoms. As an example, the alloy Fe-C is representative of an interstitial solid solution, whereas the alloy Cu-Sn (bronze) is substitutional.

The composition (or concentration) of an alloy can be expressed in terms of its constituent elements: weight (or mass) percent or atom percent. The basis of weight percent (wt%) is the weight of an element relative to the total alloy weight, whereas in the case of atom percent (at%) the calculation is based on the number of moles of an element in relation to the total moles of the elements in the alloy.

For some systems, discrete intermediate compounds rather than solid solutions may be obtained and these compounds have distinct stoichiometry and structure; for metal–metal systems, they are called intermetallic compounds. Intermetallic compounds exists as homogeneous, composite substances and differ discontinuously in structure from that of the constituent metals. Their properties cannot be transformed continuously into those of their constituents by changes of composition alone, and they form distinct crystalline species separated by phase boundaries from their metallic components and mixed crystals of these components. Although intermetallic compounds are noticed by chemical formulas, their composition is never well established on the sole basis of analytical data but can fluctuate in a very narrow range around the values indicated by the corresponding formula (Al3Mg4; Mg2Si; CuZn, etc.).

(18)

3.3.1 Phases and phase diagrams

A phase can be defined as a homogeneous portion of a system that has uniform physical and chemical characteristics. In this respect, a single-phase system is termed homogeneous. Systems composed of two or more phases are termed mixtures or heterogeneous systems. Usually, the properties combination of the multiphase system is different from, and more attractive than, the ones of the individual phases [Callister 2009].

A phase diagram represents the limits of stability among the phases composing the alloy in its chemical equilibrium system, with respect to variables such as composition and temperature. For an alloy of specified composition, at a defined temperature, the analysis of the phase diagram allows to determine structural information at ideal equilibrium condition, i.e. following an infinitely slow relaxation process.

When two metals are mixed together, three different cases can occur. The first one is a solid alloy characterized by a complete solubility of the components. An example of this type of perfect miscible elements in a single phase is the silver-gold alloy.

The second possibility is that the solid alloy shows only partial solubility of the metals in each other. One example is represented by the silver-copper pair. In this case, three principal types of phase diagrams can arise. The most common is the eutectic type, the second is the eutectoid, and the third is the peritectoid.

An eutectic system has a single chemical composition, which solidifies at a lower temperature than single constituents. When an eutectic reaction occurs, a liquid phase transforms isothermally and reversibly into a two-phase microstructure resulting from the solidification of the liquid having the eutectic composition.

On the other hand, the decomposition from a solid solution into two finely dispersed solid phases creates a structure called an eutectoid of two new intimately mixed solid phases. Finally, at the peritectoid composition, an isothermal reversible reaction of a liquid phase and a solid phase occurs to form a second solid phase during cooling.

The third case is the complete immiscibility of the two metals. As the temperature falls from the melt of these mutually insoluble metals, one of them will be precipitated, usually as globules of one phase in grains of the higher melting point metal. Examples of these microstructure are shown by the alloys of copper and lead, zinc and lead, iron and copper [Scott 1991].

A general dissertation about phase diagrams can be found in classical metallurgy books [Callister 2009, Scott 1991]. Here I will focus on the iron-carbide system, which is of interest for the purposes of the present study.

The thermodynamic equilibrium solid phases found on the iron–iron carbide phase diagram (Fig. 3.2-a) are ferrite (bcc), austenite (fcc), and the intermetallic compound iron carbide or cementite (Fe3C).

On the basis of their composition, ferrous alloys fall into three classifications:  Wrought Irons (C<0.25 wt%)

 Steels (0.25 wt% < C < 2.14 wt%) hypoeutectoid steel (0.2-0.77% C) eutectoid steel (0.77% C)

hypereutectoid steel (0.77-2.14% C)  Cast irons (C>2.14 wt% C)5.3_{[Callister 2009].}

5.3_{As a general deﬁnition, a steel is an alloy of iron, carbon (under 2% C), and other alloying elements that is}

(19)

Moreover, three invariant reaction can be defined:

 eutectoid (727°C at C 0.77 wt%), the most important one from a technological point of view.

 eutectic (1147°C at C 4.3 wt%)  perieutectoid (1495°C at C 0.17 wt%)

The development of microstructure for many iron–carbon alloys and steels depends on a eutectoid reaction in which the austenite phase (composition 0.76 wt% C) transforms isothermally, at 727 °C, to ferrite (0.022 wt% C) and cementite (6.67 wt% C) (Fig. 3.2-b). The microstructural product of an iron–carbon alloy of eutectoid composition is pearlite, a microconstituent consisting of alternating layers of ferrite and cementite (Fig. 3.2-b). The microstructures of alloys having carbon contents less than the eutectoid (i.e., hypoeutectoid alloys) are composed of a proeutectoid ferrite phase in addition to pearlite (Fig. 3.2-c).

Pearlite and proeutectoid cementite constitute the microconstituents for hypereutectoid alloys—those with carbon contents in excess of the eutectoid composition (Fig. 3.2-d).

a b

c

d

Figure 3.2: The iron–iron carbide phase diagram a and a schematic representations of the microstructures for

an iron–carbon alloy of eutectoid composition b, hypoeutectoid composition c and hypereutectoid composition d [Callister 2009]

3.3.2 Continuous cooling transformation

Phase diagrams provide no information about the time-dependence of transformation progress. However, the time element is incorporated into isothermal transformation diagrams, plots of temperature versus time (usually on a logarithm scale), and similar. They are generated from a series of plots of percentage transformation versus the logarithm of time taken over a range of temperatures.

Conversely isothermal heat treatments are not the most practical to conduct because the process can be considered valid only for conditions of constant temperature in which an

iron, carbon (over 2% C), and other elements and is not normally capable of being hot and/or cold deformed. Various types of steels and cast irons are classiﬁed and deﬁned [Bramfitt 2002].

(20)

alloy is rapidly cooled to, and maintained at, an elevated temperature from a higher temperature above the eutectoid. On the contrary, most heat treatments for metals involve the continuous and constant changing in temperature. For continuous cooling, the time required for a reaction to begin, and to end, is delayed. A plot containing such modified beginning and ending reaction curves is termed a continuous cooling transformation diagram (Fig. 3.3-a). These diagrams make possible the prediction of microstructural products for specified heat treatments. This feature was demonstrated for alloys of iron and carbon, such as perlite, bainite, spheroidite and martensite. Operatively, the rate of cooling can be influenced by the operative method, namely cooling the iron either in air, oil, or water.

In particular, the martensite phase is an important microstructural product. This is an iron– carbon solid solution with a body-centered tetragonal (bct) crystal structure. Martensite is a metastable single-phase structure, supersaturated in carbon, product of a diffusionless (athermal) transformation from austenitized iron–carbon alloys (Fig. 3.3-b). It appears by rapidly quenching austenite to a sufficiently low temperature so as to prevent carbon diffusion and the formation of pearlite and/or bainite.

a b

Figure 3.3: On the left, the complete isothermal transformation diagram for an iron–carbon alloy of eutectoid

composition: A, austenite; B, bainite; M, martensite; P,pearlite. On the right, a photomicrograph at 1220x showing the martensitic microstructure. The needleshaped grains are the martensite phase, and the white regions are austenite that failed to transform during the rapid quenching [Callister 2009].

3.4 The microstructure of metals

There are two possible basic steps of manipulating metals or alloys: the first one is casting. After that, the metal can be left, as-cast, or subsequently worked.

3.4.1 Casting

The properties of the final metal product depend strongly on the quality of the original cast metal, which is therefore very important. The types of microstructure that can arise during the casting and cooling of a melt in a mold, regardless of the exact nature of the technology involved, influence the following properties of the cast metal. They are:

(21)

 The segregation of the alloying elements.  _{The microstructure (grain size, phases).}  The soundness (porosity in the metal).

Segregation is the distribution of the solute product within the grains by solidification. Dendritic growth is actually one form of segregation that can occur during casting. It is a segregation phenomenon that often arises in impure metals, or alloys, because one of the constituents usually has a different melting point. The faster is the rate of cooling, the smaller are the dendrites.

The other principal types of segregation are normal segregation and inverse segregation. Normal segregation occurs when the lower melting point constituent is concentrated towards the inner part of the mold, while inverse segregation can push the alloying element to the exterior of the surface of the mold.

Occasionally metal can be relatively free of impurities, on slow cooling no dendrites appears, and an equi-axial, hexagonal grain structure forms. An equilibrium structure of equi-axed hexagonal crystal structure, in which all the grains are roughly of the same size, randomly oriented, and roughly hexagonal in section, can be obtained by extensive annealing the original dendritic structure.

Cast metals often display characteristic holes, or porosity, which can be due to dissolved gases in the melt, or to interdendritic holes and channels that have not been kept filled with metal during solidification. As the metal cools, and the original gases dissolve, air can

penetrate these cavities and create reactions with the metal itself to form oxides [Scott 1991].

3.5 Working

The mechanical properties of a metal are described in terms of stiffness, strength, hardness, ductility, and toughness. The behavior of the material reflects the relationship between its response, or deformation to an applied load. In particular, elastic deformation arises in the case of a temporary deformation that is totally recovered upon release of the applied stress. Plastic deformation can occurs if, after release of the applied load, the deformation is permanent or non fully recoverable. It is accompanied by permanent atomic displacements [Callister 2009].

3.5.1 Plastic deformation in polycrystalline materials

3.5.1.1 Slip system

Plastic deformation corresponds to the motion of dislocations, linear crystalline defects in correspondence with atomic misalignment, in response to an applied shear stress. Edge, screw, and mixed dislocations are possible as the result of dislocation motion.

Dislocations move with different degree of ease in relation to the crystallographic planes of atoms and the crystallographic directions involved. Usually dislocation motion occurs along specific directions in a preferred plane called the slip plane; the direction of movement is called the slip direction. This combination of the slip plane and the slip direction is termed the slip system.

The crystal structure of the metal and the atomic distortion that accompanies the motion of a dislocation at a minimum position defines the slip system. Ordinarily the slip plane is the plane that has the most dense atomic packing for a particular crystal structure, while the

(22)

slip direction corresponds to the direction, in this plane, that is most closely packed with atoms.

The different possible combinations of slip planes and directions generate several independent slip systems for a particular crystal structure.

In general, fcc and hcp crystal structures are characterized by a relatively large number of different slip systems (at least 12), being the plastic deformation possible along the entire system and reflecting in a quite ductile behavior [Callister 2009].

These crystal imperfection, the dislocations, enable deformation to take place at lower applied stress than it would be possible if the lattice structure was perfect, give rise to the characteristic malleability of metals [Scott 1991].

5.5.1.2 Mechanisms of strengthening in metals

Since the macroscopic plastic deformation corresponds to the motion of large numbers of dislocations, the ability of a metal to plastic deformation depends on the ability of dislocations to move.

The mechanical strength of a metal may be enhanced by reducing the mobility of dislocations, being hardness and strength related to the ease with which plastic deformation can occur. On the other hand, the more unconstrained is the dislocation motion, the greater is the facility with which a metal may deform, consequently becoming softer and weaker. Virtually, all strengthening techniques restrict or hinder dislocation motions making the material harder and stronger. The main strengthening mechanisms for metallic materials are hardening, grain size reduction, and solid-solution strengthening.

Strain hardening is just the enhancement in strength and consequent decrease of ductility of a metal as it is plastically deformed. The process is also known as cold-working because the temperature at which deformation takes place is “cold” in comparison to the melting temperature of the metal itself. The treatment usually consists of rolling, hammering, or drawing at room temperatures where the hardness and tensile strength are increased with the amount of cold-work, but the ductility and impact strength are reduced.

During plastic deformation, the dislocation density increases, the average distance between adjacent dislocations decreases, and dislocation mobility becomes more restricted because dislocation–dislocation strain field interactions, are, on average, repulsive.

The size of the grains, or average grain diameter, in a polycrystalline metal influences the mechanical properties. During plastic deformation, slip or dislocation motion must take place across the common boundary among adjacent grains normally characterized by different crystallographic orientations. The grain boundary acts as a barrier to dislocation motion because the two grains are of different orientations, and the atomic disorder within a grain boundary region will result in a discontinuity of slip planes from one grain into the other. More grains produce a higher strength and, therefore, a fine-grained material is harder and stronger than a coarse-grained one, since the former has a larger total grain boundary area to block dislocation motion.

In a solid solution, the strengthening of the metals is also obtained by alloying impurity atoms which go into either substitutional or interstitial positions. Increasing the concentration of the impurities result in an increase in tensile and yield strengths because impurity atoms impose lattice strain on the surrounding host atoms. This results in an interaction between dislocations and these impurities, which restrict the dislocation movement.

As an example of the various microstructures that may be produced for a given steel alloy, martensite is the hardest and strongest one and, in addition, the most brittle. Its strength and hardness are not related to microstructure. Rather, these properties are attributed to the

(23)

effectiveness of the interstitial carbon atoms in hindering dislocation motion, and to the relatively few slip systems for the bct structure.

Concerning iron and steel, carburizing is a process for case (surface) hardening obtained by absorbtion of carbon freed when the metal is heated in the presence of a carbon bearing material, such as charcoal or carbon monoxide. Because carbon is almost insoluble into ferrite phase (0.02% maximum at 730°C) the metal is heated and moved to the austenitic phase, in order to solubilize a higher percentage of carbon in it (up to 2.1% at 1150°C). Heating longer (T> 750°C) in a closed environment of saturated carbon, the solid state diffusion of carbon in the austenitic matrix occurs.

The metal can be deformed by cold-working until it is too brittle to be worked any further. If further shaping or hammering is needed, then the metal must be annealed3.4_{in order to}

restore ductility and malleability. The properties and structures may revert back by appropriate heat treatment (sometimes termed as annealing treatment). During this process of heat-treatment carried out on a metal or alloy, usually to soften the material allowing further deformation, the material is exposed to an elevated temperature for an extended time period and then slowly cooled. The mechanical properties of the material that were changed during the cold-work process, are restored and the material becomes softer, weaker and more ductile. The extent of annealing depends both on temperature and time. Together with temperature, time is another important parameter to be considered: a too long annealing can lead to grain growth and weakening of the structure, whereas a too short annealing may not be able to sufficiently eliminate heterogeneity and residual stresses.

The ductility and toughness of martensite may be enhanced and these internal stresses relieved by a heat treatment known as tempering. Tempering is accomplished by heating a martensitic steel to a temperature below the eutectoid for a specified time period. Normally, tempering is carried out at temperatures between 250 and 650°C.

3.6 Stress and strain

Cold-working a polycrystalline metal produces microstructural and property changes and some fraction of the energy expended in deformation is stored in the metal as strain energy, which is associated with tensile, compressive, and shear zones around the newly created dislocations. In defined environmental conditions, the mechanical properties of materials are ascertained by applying some load in tensile, compressive, or shear directions, constant with time or fluctuating continuously.

In the case of loading in tension and compression, stress is defined as the instantaneous load divided by the original specimen cross-sectional area.

The response of a metal to an applied force is a deformation, contracting when subjected to compression or stretching when under strain. The new configuration, assumed under the action of external forces, corresponds to a variation of the internal forces in the solid in order to balance the external ones.

As stated by Hooke’s law, the degree to which a structure deforms or strains depends on the magnitude of the applied stress. The strain is expressed as the change in length (in the direction of load application) divided by the original length.

When most materials are deformed elastically, stress and strain are proportional. For tensile and compressive loading, the slope of the linear elastic region of the stress–strain curve is the modulus of elasticity (E), which can be seen as stiffness, or as material’s

(24)

resistance to elastic deformation. The elastic deformation is non-permanent and is totally recovered upon release of an applied stress.

If a critical value of applied stress is overtaken (yield point), a permanent deformation takes place in the material, which is not recovered after the releasing of the load. This type of deformation is called plastic deformation and it is accompanied by permanent atomic displacements. An important curve for the understanding of the response of the material under a load is the so-called strain-stress curve, which is experimentally constructed from the load elongation measurements (Fig. 3.4).

The shape and magnitude of the stress-strain curve of a metal will depend on its composition, heat treatment, prior history of plastic deformation, and the strain rate, temperature, and state of stress imposed.

When stress continues in the plastic regime, the stress-strain passes through a maximum, called the tensile strength, and then falls as the material starts to develop a neck and it finally breaks at the fracture point.

Heterogeneous plastic deformations, thermal contractions and phase transformations induced by the manufacturing process can produce permanent deformation in the metal. When plastic deformation occurs, residual internal stresses forms, resulting in internal elastic strain of the material. The tension or compression remaining in the bulk without application of the external load is known as residual stress.

The residual stresses generated mechanically generally derived from non-uniform plastic deformation between the volume and the outer surface of the material; those from thermal processes are related to a non-uniform heating and cooling treatment; finally those associated with changes of volume or phase during chemical reactions such as precipitation or transformation of second phases.

Depending on the scale at which the matter is analyzed, three kinds of residual stresses are usually defined: the macro stresses, or stresses of first type, over a few grains; the stresses of second type over one particular grain; the stresses of third type across sub-microscopic areas, say several atomic distances within a grain. The stresses of second and third type are also called micro stresses. Residual stress is produced by heterogeneous plastic deformations, thermal contractions and phase transformations induced by the manufacturing process.

3.7 Texture

The crystallographic orientation refers to how the atomic planes in a volume of crystal are positioned relative to a fixed reference. This characteristic applies to all crystalline solids . Figure 3.4: a) Typical stress–strain behavior for a

metal showing elastic and plastic deformations, the proportional limit P, and the yield strength as determined using the 0.002 strain offset method [Callister 2009].

(25)

Almost all of these materials are polycrystalline and their component units are referred to as crystals or ‘grains’.

Grain orientations in polycrystals (whether naturally occurring or artificially produced) are rarely randomly distributed. In most materials, an orientation pattern is present and the occurrence of specific orientations is caused, firstly, during crystallisation from a melt or amorphous solid state and, subsequently, by further thermo-mechanical processes. This tendency is known as preferred orientation or, more concisely, texture. The importance and significance of texture to materials lies in the fact that many material properties are texture-specific [Randle and Engler 2009].

The knowledge and the measurement of texture in metals and in metal artefacts is very important. The orientation distribution is the result of the manufacturing or deformation process and thus texture contains information on the production techniques of a work piece [Scott 1991, Kocks et al. 1998, Nicodemi 1994]. Well-defined textures are produced by specific conditions during primary crystallisation from a melt, and by thermal and mechanical treatments of the cast such as annealing, drawing, hammering and rolling [Hatherly 1979].

(26)

Chapter 4

(27)

4 Methodology

4.1 The Interaction of Neutrons with Matter

The neutron is a subatomic particle having no electric charge and a mass of 1.675 10-27 kg (1,839 times that of the electron). While bound neutrons are stable in an atomic nucleus, a free neutron decays into a proton, electron, and antineutrino with a mean lifetime of approximately 900 seconds. Neutron has no charge, it has a spin (½), and a magnetic dipole moment (-1.913 µn).

The neutron can be described as a classical particle with mass m. However, according to Quantum Mechanics, it can be associated with its de Broglie wavelength λ. According to their kinetic energy, neutrons are classified according to Table 1.

Neutrons Energy range Wavelength [Å] Velocity [m/s] ultra cold ≤ 300 neV ≥ 500 ≤ 8

very cold 300 neV - 0.12 meV 52.2 – 26.1 7.5 – 152

cold 0.12 meV - 12 meV 26.1 – 2.6 152 – 1515

thermal 12 meV - 100 meV 2.6 - 0.9 1515 - 4374

epithermal 100 meV - 1eV 0.9 - 0.28 4374 - 13.8 103

intermediate 1eV - 0.8MeV

fast > 0.8MeV

Table 4.1:Classification of neutron basing on their energy [http://www.psi.ch/niag/neutron-physics].

Differently from X-rays, and photons in general, which interact with the atomic charge distributions, neutrons interact directly with nuclei according to the so-called Fermi pseudo-potential. As a consequence, while the total X-ray cross-section grows monotonically as the square of the atomic number, the neutron cross-section appears as a much less predictable function, though not too much varying, among various nuclei. As a consequence, using neutrons we are able to distinguish between different isotopes and neighbouring atoms in the periodic table of the elements.

Moreover, thanks also to the overwhelming neutron cross-section of hydrogen, with respect to the average value of all the others, the presence of even a small quantity of hydrogen (or a hydrogen-based material) can be detected behind a thick metal wall.

The interaction of neutrons with dense matter is weak, (though not negligible) making them a highly penetrating probe. On the contrary, charged particles like electrons, due to their high charge/mass ratio, are characterized by a poor penetration power below the open surface of a sample. In this respect, thermal neutrons outperform other available probes, with a penetration range of the order of several cm, even in dense materials [Squires 1996].

Thanks to the generally weak interaction with the nuclei, thermal neutrons represent a truly invasive, non-destructive probe that generally leaves the samples unperturbed, apart from a weak radioactive activation that expires, typically, in one-week time or less [Siano et al. 2002, 2003]. Of course there are exceptions to this rule, e.g. strongly absorbing nuclei, but these cases are well known and samples can be tested in advance to avoid such

Figura 4.1: Neutron interactions

with matter [http://www.psi.ch/ niag/neutron-interaction-with-matter#].

(28)

problems.

The average wavelength associated to a thermal neutron beam (E=25 meV) is around 1.8 Å, i.e. similar to the interatomic spacings. This feature, makes thermal neutrons an ideal probe for diffraction measurements.

When a neutron approaches a single nucleus, different events are possible:

a) Transmission: the neutron passes the atom without variation in its energy and direction. b) Absorption: the neutron is absorbed transferring its energy and momentum to the nucleus, which undergoes a nuclear reaction, with the formation of a different nuclide and the possible consequent nuclear decay.

c) Scattering: the energy and direction of the neutron change. The scattering can be elastic with no energy transfer from probe to sample, or inelastic. The momentum transfer is the main variable entering in diffraction experiments.

The microscopic cross-section σ is used to express the probability, for a travelling neutron with a certain velocity, to interact with target nucleus [Rinard 1991; Markgraf and Matfield 1992]. The total microscopic cross-section σT sums up the probabilities of the possible interactions, mainly scattering σs and absorption σa [Beckurts and Wirtz 1964].

a s

T σ σ

σ = + (4.1)

Of these, the absorption cross section is generally linear in the wavelength value, while the scattering cross section can be further divided into coherent and incoherent components. It is appropriate to consider an attenuation coefficient when considering the attenuation of a neutron beam by a sample. This is designed by Σ and determines the power of removal of the neutrons from the selected beam due to both, absorption and scattering processes. In a simplified picture, the beam attenuation trough a flat slice of a homogeneous material is given by the Beer’s law:

Nd=N0 exp[-ρ σT d] (4.2)

where ρ is the atomic number density, σΤ is the total cross section, and d is the thickness of the selected slice.

Neutrons represent an ideal tool to probe the microscopic properties of bulk dense materials [Squires 1996, Sears 1992] and can be used to characterize the microscopic structure (at atomic level) of matter. For metal samples, neutron techniques are used to determine the qualitative and quantitative presence of different phases, as well as the presence and distribution of textures and residual stresses at the atomic level. From this wealth of data it is possible to obtain information on the conservation status of the artefact, as well as the smelting and smithing procedures, through identification of some peculiar signatures related to these processes [Lehmann and Hartmann 2010].

Hereafter, I will focus on Neutron Imaging and Neutron Diffraction methods adopted in the present study.

4.2 Neutron Imaging Methods

There are several possibilities how neutron imaging can be used for non-destructive evaluation. Basically, one can record single radiographic images or, by collecting several transmission images and applying suitable reconstruction programmes, build a 3D tomographic view of rigid objects.

In facts, the neutron transmission through material layers depends on the specific attenuation properties of that material. Neutron imaging is based on the degree to which an object, within the beam path, attenuates the incoming radiation according to the sample’s removal cross section. The result is a shadow image of the object yielding information on

Development of innovative methodologies for non-invasive characterization of metal artefacts of archaeological, historical, and industrial interest, through neutron diffraction and neutron imaging techniques

DOTTORATO DI RICERCA IN

Scienza per la Conservazione dei Beni Culturali

Abstract

Contents

1

Introduction

Chapter 2

2

Ancient metallurgy

Chapter 3

3

Materials

3.1

Metals

3.2

Crystal structure

3.3

Alloys

3.4

The microstructure of metals

3.5

Working

3.6

Stress and strain

3.7

Texture

Chapter 4

4

Methodology

4.1 The Interaction of Neutrons with Matter

4.2 Neutron Imaging Methods